OBJECTIVE To estimate the incidence of medication errors in Spanish intensive care units. METHODS Post hoc study of the SYREC trial. A longitudinal observational study carried out during 24 hours in patients admitted to the ICU. METHODS Spanish intensive care units. METHODS Patients admitted to the intensive care unit participating in the SYREC during the period of study. METHODS Risk, individual risk, and rate of medication errors. RESULTS The final study sample consisted of 1017 patients from 79 intensive care units; 591 (58%) were affected by one or more incidents. Of these, 253 (43%) had at least one medication-related incident. The total number of incidents reported was 1424, of which 350 (25%) were medication errors. The risk of suffering at least one incident was 22% (IQR: 8-50%) while the individual risk was 21% (IQR: 8-42%). The medication error rate was 1.13 medication errors per 100 patient-days of stay. Most incidents occurred in the prescription (34%) and administration (28%) phases, 16% resulted in patient harm, and 82% were considered "totally avoidable". CONCLUSIONS Medication errors are among the most frequent types of incidents in critically ill patients, and are more common in the prescription and administration stages. Although most such incidents have no clinical consequences, a significant percentage prove harmful for the patient, and a large proportion are avoidable.